A collaborative filtering algorithm to control the quality variance in an environment-sensitive process manufacturing industry

Abstract

Optimizing the process parameters is recognized as one of the most important steps to reduce the manufacturing variance. In this paper, we proposed a collaborative filtering (CF) algorithm in which the process parameters are optimized referring to the most similar well-controlled historical records in an environment-sensitive process manufacturing. A performance metrics was proposed and numerical studies were conducted to demonstrate the algorithm's validity. The numerical results show that using the Euclidean similarity can achieve a better performance. The findings suggest that the CF algorithm can be effectively applied to not only commerce domain but also manufacturing domain and offer another applicable option to accomplish the precise manufacturing.